Author(s)

G. Chandroth, A.J.C. Sharkey & N. E. Sharkey

Abstract

The vibration signatures of a normally aspirated diesel engine contain valu- able information on the health of the combustion chamber components. It could be used to detect incipient faults in the engine. Several commonly occurring faults were induced in a 4-stroke diesel engine and the ensuing vi- bration signals recorded. Three different feature extraction techniques: Do- main expertise, wavelet analysis and principal components analysis (PGA), were compared to evaluate the effectiveness of each method. The extracted features were then used to develop artificial neural nets based fault dia- gnostic systems. It was found that the best results were obtained with the wavelets based feature extraction technique. However, each of the three sys- tems was shown to exhibit

Keywords